基于无人机影像的危岩体识别及公路地震风险研究

我国西南山区地质条件复杂，且地震频发。地震除引发滑坡、泥石流外，也会导致大量崩塌落石灾害，对基础设施造成严重破坏。以位于程海断裂带沿线、地震发生风险较高且危岩带分布典型的八代村为研究对象，分析地震作用下危岩体的影响范围，并评估山脚高速公路的受灾风险。首先基于无人机影像，利用图像识别技术快速获取危岩体位置及尺寸参数，与人工识别结果对比发现识别准确率达76.2%;然后将危岩体参数代入二维数值模拟软件Rocfall,并结合地震能量计算公式，计算地震作用下危岩体运动距离。结果表明，公路与危岩体影响区最近距离仅57 m,需要在坡脚段设置挡墙等防护措施。

Recognition of dangerous rock mass and seismic risk analysis of highway based on UAV images

In mountainous areas of southwest China, the geological conditions are complex and earthquakes occur frequently. In addition to landslides and debris flows, earthquakes can also lead to a large number of collapse and rockfall disasters, causing serious damage to infrastructure. Badai Village, which was located along the Chenghai fault zone with high risk of earthquake and typical distribution of dangerous rock mass, was taken as the research object in this paper. The influence area of dangerous rock mass under an assumed earthquake was analyzed, and the disaster risk of the expressway at the foot of mountain was evaluated. Based on unmanned aerial vehicle (UAV) images, the position and dimension parameters of the dangerous rock mass were quickly obtained by using the image recognition technique, and the recognition accuracy reached 76.2%. Then, the parameters of dangerous rock mass were substituted into the two-dimensional numerical simulation software Rocfall, and combined with the seismic energy calculation formula, the movement distance of dangerous rock mass under seismic action was calculated. The calculation results show that the nearest distance between the highway and the impact area is only 57 meters, so protective measures such as retaining walls should be built at the foot of slope.